I’m encountering power related issue on my uConsole (CM5) that initially appeared to be isolated to the keyboard, but may point to a broader power delivery problem.
Primary symptom:
The built-in keyboard disconnects when plugging any device to the usb port and re-enumerates as Leaflabs Maple DFU (1eaf:0003), at which point it no longer functions until rebooted (reboot solves the problem only while running on the battery).
Sometimes it also disconnects while trying to flash ESP32 based board using the USB.
The issue persists even when a high-power USB-C supply is connected, which suggests the bottleneck may be the 5V boost stage or downstream distribution rather than batteries.
What is suspected:
The CM5’s higher sustained and peak current draw compared to CM4 could be stressing the stock boost converter and PCB distribution.
Objective of this post:
I’d like to gather input from those familiar with the uConsole’s power architecture to determine:
Whether replacing the existing 5V boost converter with a more powerful module is feasible.
Whether adding bulk capacitance at strategic points would be an effective mitigation.
If there are known weak points or choke points in the current power path worth reinforcing.
I’m interested in solutions that address the root cause rather than just the keyboard symptom (if it would be feasible).
The USB port disconnecting the keyboard happens as often on the cm4 and doesn’t seem to be directly related to power draw issues as USB ports installed in the expansion slot are unaffected.
To preface, this is on a CM4, so my problems will inevitably be different. My expansion port has HackerGadgets’ SDR card.
I’m trying to find ways to easily reproduce this problem on my system. I’ve found that plugging in all of an external keyboard (HHKB Professional Classic with power saving enabled), a mouse (I had been using a Logitech M110 but am now using a Logitech wireless receiver), and a USB flash drive, and then mounting the drive, writing to it, and unmounting it, and then unplugging everything at once can cause the built-in keyboard/trackball to disconnect. Most recently when I performed this, the built-in HIDs reconnected and are now useable again, but previously I had seen them fail to reconnect. PS: also, with the mouse and keyboard plugged in, simply adding the flash drive has caused disconnection.
Others who are smarter than me seem to have concluded this is not a protocol issue, but if it would be useful, I can try to get a PCAP of the USB traffic on my machine around the power failure - please let me know, as I don’t want to waste the kilobytes of site storage otherwise.
I don’t think I would notice the speaker symptom because I almost always have them muted. I do notice the speakers pop when powering the console on or off, and the first sign that it’s powering on is that the LCD backlight flickers.
I’m not sure any. I didn’t do any rigorous testing. But my impression is that it made it slightly less likely to drop out. Part of the problem with intermittent issues is it’s hard to establish a pattern. Especially since I pretty much only use a USB expansion port these days.
First of all, CM5 is power-hungry.it needs 5Amps, I recommend to downclock a little bit.
You can use better quality cells 18650 with higher discharge current rating. 10A+. I have sony(murata) vtc5A cells in my uconsole with cm5 board, and for what i use I cant reproduce this problem.
Yeah, now I’m regularly experiencing the keyboard disconnecting. It’s either a reboot or popping the front off and replacing it to fix it, and it happens when I plug more than a couple things in.
When plugging in any USB device (except ones that have their own power), I have keyboard disconnection issue with fully charged 4000mAh Nitecore batteries even while uConsole being powered by 3A USB-C at the same time.
Could this be related to the bypass capacitor after the TPS2553?
It sounds like we should have 120uF per port, and we only have a single 100uF in the drawing for the Mainboard V3.14 V5.
Edit: The USB power switch (TPS2553) has a current resistor of 24.9kohms; so the nominal current limit is ~1amp. I can’t imagine that’s tripping from 1-USB A device and the built-in keyboard.
I just recently got my uConsole. If I can recreate your issue, I’ll replace that capacitor and see what happens.
Something small, like a mouse receiver, didn’t cause the error. Something a little bigger, a USB drive with an LED, did. (I used dmesg -w to check, I’m still using stock keyboard firmware)
I replaced C513 with a 220uF tantalum capacitor (that’s just what I had), and I’m not seeing the issue when using my USB drive anymore. I’ve also tried a few other devices, and nothing.
Edit: I originally added a 220uF capacitor for a total of 320uF. I didn’t like the install so I just replaced the capacitor instead. Both ways seemed to work.
Just in case if someone is interested in such mod:
Type: type B / 3528 / 3.5x2.8 mm
Capacitance: 220uF
Voltage: 10v(but I could only find 6.3V for a reasonable price)
